University of Virginia Health System

Email Address: jwf8x@virginia.edu

This project was designed to make available, chemical-tag technology for determining the relative ratios of proteins present in different samples to the Mellon investigators and to the University of Virginia at large. This technology, also called isotope-coupled affinity tags (ICAT), has been touted as an alternative to comparative 2-D PAGE (2-dimensional polyacrylamide gel electrophoresis) for assessing the relative amounts of different proteins in separate samples. These reagents may have potential application for several research programs involved in proteomics within the Mellon Institute (Theodorescu, Weber, Parsons etc.).

Initial experiments performed on BSA (bovine serum albumin) or a mixture of four standard proteins at defined ratios were successful in determining reaction conditions. Results indicate that the error for such experiments will be +/-30 percent. Problems arose in the ratio determination, targeting of ions for sequence analysis and peptide identification. Unfortunately, current software available in the Biomolecular Research Facility is not sufficient to handle these problems, resulting in missed identifications and large amounts of manual work determining protein ratios. It is possible that the necessary software has not yet been developed.

A whole cell mixture was used as a second test case. While the chemistry works, the above problems (magnified now by the large numbers of proteins present) make the experiment somewhat impractical without far better software and sample fractionation (subcellular, protein, and peptide fractionation steps are all probably needed). The Theodorescu laboratory has submitted samples for complex mixture analysis and, in conjunction with the Biomolecular Research Facility, continues to work on subcellular fractionation for possible ICAT work in the future.

In summary, using Mellon Institute support, the project has assessed and brought on-line ICAT technology for Mellon investigators. Technological limitations have been delineated, as have experimental situations where it is well suited. Although it is clearly not the final answer to determining the ratio of specific proteins among samples, it can give an approximation that will be satisfactory for some experimental paradigms.